Genstat: The Complete Data Analysis Solution

Using Genstat in the ‘sustainable intensification’ of food production

I have been an avid supporter of Genstat since starting my PhD in 1985 (gulp!). As an applied plant biologist, it is perhaps unsurprising that I found procedures that had been developed with colleagues at Rothamsted Research to be highly relevant to the analysis of my own designed experiments. During my thirty-five years in research, teaching and management at higher education institutions, I have continued to rely on Genstat as my core statistical programme both as an investigator and an educator. My own research is in the area of crop science, and in the last decade or so, justified (to research funders at least) on the basis of the need to ‘sustainably intensify’ our food production \[1\]. For many the phrase is an imperative, i.e. we have to use our land more efficiently and it will have to keep on being able to do this; for others the phrase is an oxymoron, e.g. if land-use efficiency means producing ever higher yields per unit area this will inevitably degrade the very ecosystems and resources on which agricultural production by future generations will depend; a third group would argue that sustainable intensification is unnecessary – if (big if) we were each satisfied with the per capita calorific availability that was the average for humans in 1960, we are already producing enough calories to feed 10 billion people. Hopefully my research has been relevant whichever ‘camp’ you may belong to. In my early days as a researcher I was primarily interested in the optimising of rates of synthetic inputs (such as fertilizers and crop protection chemicals) to achieve yield and grain quality targets in wheat, to improve economic and environmental sustainability. Experiments were usually complete multi-factorial designs (genetic x management), replicated in randomised blocks usually with some split plotting. Given Genstat’s origins, it is unsurprising that the treatment structure and block structure routines allow a very large degree of flexibility to cope admirably with all types of designed field experiment. Of particular use to me was the ability to split treatment effects into polynomial contrasts within ANOVA \[2, 3, 4\]. ‘Life is not a polynomial’ but this approach provides an initial, statistically sensitive test as to whether there is a response to increasing rates of an input, and some idea as to the complexity of that response. Guided by the ANOVA, the FITCURVE routine provides a straightforward step for exploring more biologically meaningful and parsimonious fits to the responses to inputs \[5,6\]. The standard curves available in Genstat are highly relevant to the applied biologist but further development and modification of responses is supported by FITNONLINEAR \[7,8\].  FITCURVE and FITNONLINEAR have also been useful in fitting responses over time. Often treatment (such as fungicide) effects on disease or green leaf area on individual days are associated with effects on the errors around the treatment means. This variance heterogeneity is easily identified in the residual plots provided by ANOVA and can be rectified by transformations. Transformed data, however, can be difficult to interpret, whereas fitting responses over time (e.g. logistic or gompertz) can generate new variables such as time scalars which are readily interpretable, and for which the residuals are often normally distributed and lacking variance heterogeneity \[7\]. Over the years the teams of PhD students and researchers I’ve been involved with have continued to be heavily reliant on ANOVA, FITCURVE and FITNONLINEAR although the responses these days are more likely to be due to climate variables (genetic x environment) \[9\]. As datasets and experiments have got bigger and less complete, residual maximum likelihood (REML) has proved more flexible \[10, 11, 12\] than ANOVA, and principal components (PCP) analysis ably condenses large variate x experimental unit matrixes \[13,14\]. The current portfolio of procedures is well matched for advances in applied biology whether involving meta-analyses, ‘big data’, statistics to underpin machine learning, and/or precision agriculture. As an educator I have taught BSc and MSc statistics and research methods modules with Genstat, and have enabled level 5, 6, & 7 students using the programme to assess class-generated data. The interface is intuitive, and it is easy to demonstrate variations in approaches to highlight the importance of, for example, replication (and the curse of pseudo-replication), blocking structures, covariance, randomisation and the like. Genstat has been the main statistical programme used by the 40 PhD students who have been gracious enough to allow me to be involved in their supervision.   1 Reaping the benefits: science and the sustainable intensification of global agriculture ISBN: 978-0-85403-784-1; 2_J. Agric. Sci_ **138**, 317-331, doi: 10.1017\\S0021859602002137; 3_J. Cereal Sci._ **37**, 295-309, doi: 10.1006/jcrs.2002.0501; 4_Plant Soil_ **360**, 93-107, doi: 10.1007/s11104-012-1203-x; 5_J. Sci. Food Agric._ **85**, 727-742, doi: 10.1002/jsfa.2025; 6_J. Agron. Crop Sci._ **200**, 36-45, doi: 10.1111/jac.12038; 7_Annals Appl. Biol._ **136**, 77-84, doi: 10.1111/j.1744-7348.2000.tb00011.x; 8_Field Crops Res._ **95**, 49-63, doi: 10.1016/j.fcr.2005.02.001; 9Frontiers Plant Science **8**:51, doi: 10.3389/fpls.2017.00051; 10_J Agric Sci_ **128**, 135-142, doi: 10.1017/S0021859696004054; 11_Agric. Forest Meteor._ **94**, 159-170, doi: 10.1016/S0168-1923(99)00020-9; 12_PLoS One_ **11**(5), e0156056, doi: 10.1371/journal.pone.0156056; 13_J. Sci. Food Agric._ 84, 227-236, doi: 10.1002/jsfa.1657; 14Euphytica **166**, 249-263, doi: 10.1007/s10681-008-9838-7 Mike Gooding FRSB (Mike has headed agriculture and been Professor of Crop Science at The University of Reading, Aberystwyth University, and the Royal Agricultural University).

KWS: supplying seeds to the farming industry

KWS is one of the world’s leading suppliers of seeds to the farming industry and it is therefore no surprise that research into plant breeding and seeds is a crucial part of their work. ## KWS value precision and security For centuries people have been breeding plants, trying to fine-tune individual plants to obtain or change a particular trait. Today plant breeders look at addressing concerns such as drought or environment tolerance, pesticide, herbicide, fungi or bacteria tolerance, improving nutritional value or increasing yields.  This, by definition, ranges from “classical” propagation techniques to the modern methods of molecular breeding and genetical modification. However, one thing is very apparent: the scientists engaged in any form of plant breeding need to be sure their data analysis is based on sound, solid and secure statistics. KWS is one of the world’s leading suppliers of seeds to the farming industry and it is therefore no surprise that research into plant breeding and seeds is a crucial part of their work. The research and development teams at KWS focus on yield, seed quality, resistance to disease and pests, enhancing the nutrient quality and improved processing capability of the plants. Recently the researchers at KWS decided to change their data analysis software portfolio by introducing Genstat. During the last several years the main emphasis was put on establishing Genstat for statistical analysis of phenomic data from their field. Genstat was chosen, because it was possible to analyse data from trials with large Alpha Designs with the REML techniques. Plus, Genstat’s REML algorithm dealt with missing values in a more superior way than the previous software. REML itself can be used to analyse models with several types of error variation (multi-level models) and to fit models to correlated data like repeated measurements. Genstat’s powerful command language streamlined the data analyses, reducing the calculation time by at least 70%. Additionally Genstat’s efficient data handling meant that far less data pre-processing was needed. Genstat allows the full integration into automated analysis pipelines, thus decreasing calculation times by 30%. This allows the flexible addition of new statistical methods into the pipelines. KWS researchers use a wide variety of different factors in their trials, using Block Designs, Alpha Designs, Lattice, Split Plot, Multi factorial Experiments. Genstat allows the researchers to easily design and analyse these and other types of experiments. Genstat also effortlessly handles large and complex data sets, which will be essential for the analysis of genomic data and their association with phenomic informations. Combine these factors with Genstat’s ability to easily export and import data from other databases and the plant breeder has very powerful analysis tools. As well as the extensive technical capabilities of Genstat, KWS chose Genstat because of its history within agricultural science and research; it was originally designed by statisticians working on agricultural research at Rothamsted Research; where Fisher, Yates and Nelder (to mention but 3) developed statistical techniques that are central to modern statistics. Genstat’s connection with Rothamsted continues today. Developments in each new version of Genstat reflect the needs of the agricultural scientist, including ANOVA, design of experiments a host of multivariate analysis techniques and linear mixed model analysis. A huge variety and complexity of data analysis is possible in Genstat, and an understanding of the requirements of experiments in this area is reflected in the terminology and thinking behind Genstat. But Genstat’s history not only shows its suitability to any form of agricultural research but also shows the stability of its performance; it has been tried and tested for over 30 years. Coupled with the power and background of Genstat, VSNi were able to provide a bespoke consultancy service to KWS to assist with their specific needs; this is possible because our developers have been working with agricultural researchers and statistics for years. The VSNi statisticians were able to talk to the researchers at KWS, in their language to find more efficient ways to run their analyses. In a subject area that demands precision and security Genstat provides the data analysis solution that removes the unknowns surrounding the research. You can trust the statistics within Genstat, because it is developed by people who know and understand the issues in agricultural research today.

Food Standards Agency: safer food for the nation

Everything we do reflects our vision of ‘Safer food for the nation’. We aim to ensure that food produced or sold in the UK is safe to eat, consumers have the information they need to make informed choices about where and what they eat and that regulation and enforcement is risk-based and focused on improving public health.  Genstat has been deployed in many important areas of FSA activity, including the following:- 1\. Exposure Assessment There are many types of chemical whose presence in food can pose some level of risk for consumers. These include environmental contaminants (e.g. lead or cadmium), as well as certain pesticides and food additives. The FSA needs to estimate the likely level of human exposure to such chemicals, either from individual foods, or from the diet as a whole. Our exposure assessment models combine occurrence data (on the level of a given contaminant in foods) with extensive human intake data (from individual food diaries). Genstat provides excellent tools for the necessary data import, indexing, manipulation and analysis and produces fast, reliable results. It has the flexibility to perform more complex modelling when required: e.g.  probabilistic assessments of risk that allow for variability and uncertainty in some of the key input variables. 2\. Risk Assessment The FSA uses Hierarchical Generalized Linear Models to model the probability of elevated levels of naturally occurring biotoxins in shellfish. These estimates are used to prioritise biotoxin sampling activity around the shores of the UK, indicating where, and at what times of year, the level of sampling should be raised, and also where and when it can be reduced. Genstat has excellent procedures for performing models of this kind and for extracting model coefficients in a form that can be used to make a range of policy-relevant inferences.